linux-kernel - Re: [PATCH v1 1/1] soc/aspeed: Add host side BMC device driver

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date: Thu, 11 Jan 2024 10:23:29 -0600
From: Ninad Palsule <ninad@...ux.ibm.com>
To: Ninad Palsule <ninad@...ux.vnet.ibm.com>, Andrew Lunn <andrew@...n.ch>
Cc: joel@....id.au, andrew@...id.au, eajames@...ux.ibm.com,
        linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
        linux-aspeed@...ts.ozlabs.org
Subject: Re: [PATCH v1 1/1] soc/aspeed: Add host side BMC device driver

Hello Andrew,

On 8/23/23 12:32, Ninad Palsule wrote:
> Hello Andrew,
>
> On 8/22/23 11:14 AM, Ninad Palsule wrote:
>> Hello Andrew,
>>
>> Thanks for the review.
>>
>> On 8/21/23 2:29 PM, Andrew Lunn wrote:
>>>> Testing:
>>>>    - This is tested on IBM rainier system with BMC. It requires BMC 
>>>> side
>>>>      BMC device driver which is available in the ASPEED's 5.15 SDK
>>>>      kernel.
>>> How relevant is that? To the host side, it just appears to be an
>>> 16550A. Is the SDK emulating an 16550A? If you where to use a
>>> different kernel, is it still guaranteed to be an 16550A? I also
>>> notice there is a mainline
>>> drivers/tty/serial/8250/8250_aspeed_vuart.c. Could that be used on the
>>> BMC? That would be a better testing target than the vendor kernel.
>>
>> This is just to indicate how I tested my code.
>>
>> Yes, aspeed chip (in this case ast2600) is compatible with 16550 UART.
>>
>> I am guessing it should work with different kernel too as 16550 
>> standard is used.
>>
>> The 8250_aspeed_vuart.c is a BMC side driver for accessing VUART over 
>> LPC bus and
>>
>> this is a host side driver to access VUART over PCIe bus.
>>
>>>> +config ASPEED_HOST_BMC_DEV
>>>> +    bool "ASPEED SoC Host BMC device driver"
>>>> +    default ARCH_ASPEED
>>>> +    select SOC_BUS
>>>> +    default ARCH_ASPEED
>>> same default twice?
>> Removed.
>>
>>>> +late_initcall(aspeed_host_bmc_device_init);
>>>> +module_exit(aspeed_host_bmc_device_exit);
>>> It looks like you can use module_pci_driver() ?
>> yes, It should work unless the late initcall is important. I will 
>> test it and see.
>
> I will not be able to use module_pci_driver() as it doesn't support 
> late initcall which is required otherwise
>
> 8250 registration fails. So I am not making this change.

Please let me know if you are fine with this.

Thanks for the review.

Regards,

Ninad